Advances in retinal imaging modalities: Challenges and opportunities

A computer-aided diagnosis system for detecting various diabetic retinopathy grades based on a hybrid deep learning technique

Diabetic retinopathy (DR) is a serious disease that may cause vision loss unawares without any alarm. Therefore, it is essential to scan and audit the DR progress continuously. In this respect, deep learning techniques achieved great success in medical image analysis. Deep convolution neural network (CNN) architectures are widely used in multi-label (ML) classification. It helps in diagnosing normal and various DR grades: mild, moderate, and severe non-proliferative DR (NPDR) and proliferative DR (PDR). DR grades are formulated by appearing multiple DR lesions simultaneously on the color retinal fundus images. Many lesion types have various features that are difficult to segment and distinguished by utilizing conventional and hand-crafted methods. Therefore, the practical solution is to utilize an effective CNN model. In this paper, we present a novel hybrid, deep learning technique, which is called E-DenseNet. We integrated EyeNet and DenseNet models based on transfer learning. We customized the traditional EyeNet by inserting the dense blocks and optimized the resulting hybrid E-DensNet model's hyperparameters. The proposed system based on the E-DenseNet model can accurately diagnose healthy and different DR grades from various small and large ML color fundus images. We trained and tested our model on four different datasets that were published from 2006 to 2019. The proposed system achieved an average accuracy (ACC), sensitivity (SEN), specificity (SPE), Dice similarity coefficient (DSC), the quadratic Kappa score (QKS), and the calculation time (T) in minutes (m) equal [Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text], 0.883, and 3.5m respectively. The experiments show promising results as compared with other systems.

Comparison of experiences and preferences following non-invasive cardiovascular risk procedures: a cross-sectional survey in participants with and without diabetes mellitus

Aims

Endothelial dysfunction is an early risk marker of cardiovascular disease in diabetes mellitus. Timely screening is important in reducing cardiovascular disease-associated morbidity and mortality. This cross-sectional study investigates the acceptability and preferability of non-invasive cardiovascular risk procedures (EndoPAT2000 system and the ECG-gated fundoscope) in participants with diabetes mellitus compared to controls.

Methods

A self-administered Likert scale-based questionnaire was completed by 106 controls and 117 participants with diabetes mellitus, identified through stratified random sampling, upon conclusion of an Australian Heart Eye sub-study conducted at Westmead Hospital, NSW, Australia from 2012 to 2014. Pearson’s χ² test, independent-samples t-test and regression analysis were performed.

Results

Study participants who responded to the questionnaire had no preference for procedures (controls: 2.4 ± 1.1 vs diabetes mellitus: 2.5 ± 0.9, p = 0.38) but had an overall more negative experience with most aspects of the ECG-gated fundoscope than the EndoPAT2000 system. Of those with diabetes mellitus, participants who provided poorer self-rated health expressed discomfort with the mydriatic drops (ß 0.27, 95%CI 0.001 - 0.54, p = 0.049) and the fundoscope’s green light filter (ß 0.27, 95%CI 0.07 - 0.47, p = 0.009), as well as maintaining still (ß 0.40, 95%CI 0.08 - 0.72, p = 0.02) and not blinking (ß 0.38, 95%CI 0.07 - 0.70, p = 0.02) during photo acquisition. These participants were also less willing to repeat the ECG-gated fundoscope procedure (ß 0.29, 95%CI 0.07 - 0.52, p = 0.01).

Conclusions

Participants with diabetes mellitus, especially with poorer self-rated health, had a more negative experience with the ECG-gated fundoscope than the EndoPAT2000 system. Difficulties experienced under examination by the ECG-gated fundoscope appear related to the procedural design, which requires amendments improving patient comfort and compliance.